MQSA Historic success becomes regulatory threat

Thirteen years after its passage by Congress, the Mammography Quality Standards Act remains a sore subject for radiologists. While breast imagers unanimously agree that the MQSA has vastly improved mammography quality, they give the rule low marks for its cost and the burdensome details of compliance.

Now a new report urging increased mammography data collection and radiologist performance tracking, as well as an expansion of the act to breast ultrasound and MRI, is sparking rebellion. Some fear that the shrinking number of breast imagers could dwindle still further under the proposals, affecting access to breast cancer screening.

The prospect of increased regulation without accompanying upgrades in mammography reimbursement is the key complaint about the new recommendations, developed by the Institute of Medicine at the request of Congress (Table 1). But radiologists consider the government bureaucrats' lack of acknowledgment of their contribution to decreased breast cancer mortality through screening insult added to injury.

When recognition that screening could have a powerful impact on early breast cancer detection emerged in the mid-1980s, radiologists rose to the occasion, said Dr. Daniel Kopans, director of breast imaging at Massachusetts General Hospital.

"Mammography is difficult, stressful work," he said. "But since screening began, the breast cancer death rate in the U.S. has dropped by 25%. Better therapies have also contributed, but the majority of that decrease is from screening. It galls me that the huge success we have had is being ignored. Instead, the message is that radiologists need to be whipped into line."

Breast imagers also chafe at the fact that no other imaging service is subject to the same level of regulation as mammography and that the oversight is seemingly without end. But all of radiology can learn from the lessons-good and bad-of the MQSA. Indeed, some predict that the MQSA is the writing on the wall for the rest of radiology.

"Mammography is on the cutting edge, because it is the first area to get regulated and to implement quality measures that are reproducible across institutions," said Dr. Etta Pisano, director of breast imaging at the University of North Carolina, Chapel Hill. "But such regulation is going to happen across all of radiology. At some point, we won't be able to do any imaging exam without proving quality outcomes measures to payers."

MQSA LEGACY

Oversight of mammography unofficially began in 1987, when the American College of Radiology launched its voluntary facility accreditation program. The action occurred at a time when mammography screening was becoming widely adopted, yet image quality, equipment, and radiation dose for the procedure varied considerably.

Despite the ACR's efforts, by 1992, only about 7200 facilities out of an estimated 11,000 had applied for accreditation, and of those, just over 4600 were accredited, according to the IOM.

At the same time, states began implementing their own laws governing mammography quality. The resulting patchwork policies and continuing quality problems culminated in congressional action, with the passage of the MQSA in late 1992. Compliance with the act became mandatory in October 1994.

Today's MQSA is based in large part on the original ACR program. Since its passage, the MQSA has remained largely unchanged through two subsequent reauthorizations, in 1998 and 2004, with the exception of the addition of digital mammography in March 2000.

The MQSA ended the use of modified x-ray devices for mammography screening by mandating dedicated equipment. It established national quality standards and certification for physicians, technologists, and physicists. Compliance with the rule means conducting and maintaining documentation of daily, weekly, quarterly, semiannual, and annual quality control tests. Sites must apply for accreditation and undergo annual inspections at their own expense.

It's difficult to overestimate the MQSA's impact on mammography quality. Screening pioneers polled by Diagnostic Imaging used terms like "horrific" and "desperately bad" to describe the state of pre-MQSA mammography. The comments of Dr. Wende Logan-Young, director of the Elizabeth Wende Breast Clinic in Rochester, NY, are typical.

"In 1976, when I opened the clinic, mammography was atrocious," she said. "MQSA was important in bringing people into line, because the word was out to patients and referrers that facilities needed to be certified and accredited."

As a result of the act, nearly 70% of facilities now pass inspection without any violations, and fewer than 2% of violations issued are for the most serious level, according to the FDA, which administers the program. The agency keeps a running tally of certified facilities and equipment, as well as mammography procedure volume (Table 2).

"Before MQSA, radiation levels varied widely, equipment was shoddy, and physicians often did not have proper training," said Senator Barbara Mikulski (D-MD), the driving force between MQSA then and now. "Since its implementation, the quality of mammography has improved. What surprises me is that mammography is the only imaging technique that has quality standards."

Sites that could, complied with the new rules. Others dropped off the rolls of mammography facilities. That left the field of breast imaging to those who were committed to providing quality services, said Dr. Carol Lee, a professor of radiology at Yale University and director of breast imaging at Kapiolani Breast Center in Honolulu.

The MQSA also raised awareness of the importance of mammography among hospital executives, Kopans said.

"Prior to MQSA, mammography was considered the short straw, done in the basement," he said. "MQSA gave mammography screening some respect, because providing it meant meeting standards and achieving a level of quality. That got the attention of hospital administration."

QUALITY AT A COST

Such strides haven't come without their costs, financial and otherwise. Complying with MQSA audit regulations and preparing for annual site visits is a time and resource drain for breast centers. Lee faces an upcoming MQSA site visit to Kapiolani Breast Center.

"I'm dreading it," she said. "They are going to come and look and make sure all of our t's are crossed and our i's dotted."

The addition of full-field digital mammography created another wrinkle in the inspection process, said Valerie Andolina, chief technologist at Elizabeth Wende Breast Clinic. Not only is complying with MQSA rules for digital units time-consuming, but the standards that exist merely mirror vendor recommendations.

"It's a huge problem," Andolina said. "We are actually teaching the MQSA inspectors about digital mammography. They only test to see if we are doing what the manufacturer recommends. They don't know if it's right or wrong."

In addition to the hassle factor, MQSA compliance represents a significant financial burden for mammography facilities. Based on data collected by the ACR, the cost of MQSA compliance is between $14 and $15.79 per mammogram (Table 3).

Compound those costs with flat-lined reimbursement and escalating malpractice claims, and it is little wonder that there is a workforce shortage in mammography.

"The profit margin is so low for mammography that it's a money loser in every radiology department," Logan-Young said. "The few radiologists who are in it are overworked and underpaid, so why would anyone want to go into it?"

Indeed, a survey of radiology residents in 10 programs across Massachusetts found a severe lack of interest in providing mammography services once training is complete (JACR 2005; 2:432-435). Of 63 residents, only one expressed interest in a breast imaging fellowship, and only 3% of the total said that they would like to spend a substantial amount of time interpreting mammograms in the future. Residents named fear of lawsuits as the most common reason behind their lack of interest.

That finding is even more alarming than the one resulting from a 2003 survey of radiology residents, in which trainees reported that they sought to spend less than 25% of their time in practice on breast imaging (Radiology 2003; 227:862-869). That number may shrink still further, as radiologists opt out of breast imaging in favor of higher revenues and less stress in other areas of radiology. Almost one in three radiologists (33 of 122) report considering discontinuing mammography interpretation on a monthly basis, according to one published study (Radiology 2005; 236:37-46). And 16.4% report a similar desire to withdraw from weekly or daily mammography reading. Both groups cite liability fears as the reason.

Malpractice concerns are only part of the equation, however. In a study evaluating the capacity of community-based mammography facilities to deliver screening and diagnostic services in the U.S., researchers suggest that strict MQSA measures may also play a role in dissuading radiologists from pursuing a breast imaging career (Radiology 2005;235:391-395). The survey of 45 facilities in three regions of the U.S.-Colorado, the Pacific Northwest, and New Hampshire-documented shortages of radiologists to deliver needed services in 44% of the sites.

The same study also found extended wait times to obtain screening and diagnostic mammograms. With the threat of expanded MQSA regulations looming, but no promise of reimbursement or malpractice relief, radiologists worry about the future of mammography practice and about the impact on access to breast cancer screening.

"There's a fine line here," said Lee, who chairs the ACR's commission on breast imaging. "MQSA caused people who weren't committed to stop offering it, but I'm worried that with increasing regulation, good people will start saying that mammography is just too much bother. It's something we as a society have to be very careful about."

PERFORMANCE PROBLEM?

Mikulski authored and championed the MQSA's most recent reauthorization in October 2004. As part of the same vote, Congress commissioned a report from the IOM on ways to improve mammography quality to be considered in 2007 when the MQSA is next up for renewal.

That report, Improving Breast Imaging Quality Standards, released in May, does address ways to reduce paperwork burdens and increase radiologist recruitment to breast imaging. It eliminates a slew of redundant tests. The IOM directs the FDA to seriously study the workforce issue and its impact on access to care, and a General Accounting Office report on that topic is also under way.

The report also addresses the malpractice problem, offering IOM support of medical liability reform. It suggests that high-performing centers and radiologists be eligible for no-fault liability.

But the main focus of the IOM and its report recommendations is on radiologists' interpretive performance in mammography. This breaks new ground for the MQSA. The existing rule sets minimums for physician education and skills, including mandating mammography reading volumes and CME every three years. It also requires sites to track outcomes data on patients referred for breast biopsy. But the IOM proposes expanding current medical auditing regulations beyond biopsy tracking.

If the recommendations are adopted, breast centers would be required to collect data that offer additional details on physician performance and patient outcomes:

- cancer detection rates;

- positive predictive values (PPV) of women who are subsequently diagnosed with breast cancer; and

- abnormal interpretation rate (women whose mammograms lead to additional imaging or biopsy).

The rationale behind the recommendation is to acquire more reliable data on physician interpretation trends and to be able to identify those who excel, said Dr. Howard Forman, an associate professor of radiology at Yale and a member of the IOM report committee.

"Under current MQSA regulations, we do not capture any reliable information about physician interpretation," he said. "With these recommendations, we may not capture all of the data that we'd like, but we can begin moving in a direction that identifies the best interpreting physicians and practices, and where the best imaging occurs."

Providing ways for radiologists to obtain feedback on their breast cancer detection performance is critical to improving breast imaging quality, said Pisano, also an IOM committee member.

In addition to upgrading mandatory data tracking regulations, the IOM recommends that Congress institute an advanced, voluntary audit. Through this program, sites would collect tumor staging information from pathology reports and patient demographic data. A central data center would store the data and provide feedback to individual physicians.

Facilities whose data show high levels of accuracy would be designated as "centers of excellence" and could be rewarded for their high-quality services through additional reimbursement.

"We need evidence if we are to identify the metrics that improve imaging interpretation," Forman said. "What better evidence than to create centers of excellence, which can conduct pilot programs that determine the types of activities that improve interpretation?"

Even published data on the impact of initiatives such as double-reading, computer-aided detection, and high-volume reading of mammograms are highly variable, Forman said.

"We found that there is insufficient evidence to support the idea that these types of activities are better, even though many of us have long assumed that they are," he said. "What we have done is make suggestions about how to get the evidence that will allow us to make these decisions in the future."

Last but certainly not least, the IOM recommends that increased reimbursement must accompany increased regulation. That's an issue that meets with agreement from organized radiology. Leaders from the ACR and the Society of Breast Imaging have already weighed in on the report, calling for additional reimbursement before any recommendations are adopted.

The IOM committee was "very cognizant" of the payment issue, Pisano said.

"Every single item in the report that involves additional activity is accompanied by a recommendation for increased payment," she said. "We don't believe that screening mammography can absorb unfunded mandates."

But others take issue with the basic idea that more regulation will improve mammography interpretation. And Kopans and others question the likelihood of a pay raise.

"What is going to happen is that Congress will pass all of the things that need to be done, but it won't address any way to provide more money to do them," Kopans said.

Moreover, it is going to be difficult for radiologists to get on board if the IOM recommendations are adopted, in part due to what Kopans calls "terrible language" in describing interpretive failings.

"No one is looking for accolades, and we know we haven't solved the problem of breast cancer," he said. "But we've made a huge dent in it. I think the IOM makes a big mistake by not acknowledging that and enlisting radiologists' support in ways to improve."

Radiologists readily agree that mammography interpretation could be better. Everyone wants to improve, Lee said.

"But we want to improve in ways that will give some results and not more paperwork, tracking, and government regulation," she said.

Variability in interpretive performance is bound to exist as long as there are facilities that perform only small numbers of mammography screens annually, said Gerald Kolb, president of Breast Health Management, a consulting firm based in Bend, OR.

"It takes a large, high-volume practice to yield high-end expertise in mammography reading," he said. "The data show that on average, there are 3.5 to 5.5 cancers per 1000 screening mammograms. For radiologists who are reading a minimum number of exams, they have the opportunity to detect at most two cancers per year-assuming they find both of them. That's not enough to become good."

But even high-volume radiologists miss cancers. In a study performed at Elizabeth Wende Breast Clinic, double-readers picked up 15% more cancers than first readers on mammography screening exams. The site performs about 75,000 screening mammograms annually.

"The range of normal for a mammogram is so tremendous that it takes years of experience to do it well," Logan-Young said. "Every one of our doctors here does nothing but read mammograms. Even then, they will not find everything that a double-reader can. That's how difficult it is."

There are ways the U.S. could learn from other countries-Sweden, for example. That country relies on interpretation performed at centralized diagnostic facilities that use the best equipment and radiologists who are trained and interested in breast disease, Kopans said. Outlying sites perform the actual screening exams, which are read by experts at the centralized facility.

"It's the most efficient and effective way to do screening," he said. "But in this country, we don't have that kind of centralization of services, or a government-run health system."

Better ongoing training is another way breast imagers could upgrade skills. Kolb suggests that instead of CME meetings, radiologists could benefit from interactive seminars. Such sessions might involve interpreting 100 consecutive cases, with 55 cancers spread throughout.

This kind of mentored review of mammograms of biopsy-proven cancers and followed benign cases, combined with supportive coursework and self-testing, has been highly effective in increasing cancer detection rates, he said.

"You get to actually see the cancers, and you are tested in real-time on how you are doing," he said. "That is meaningful."

THE NEXT FRONTIER: BIQSA?

The IOM report's release is just the first stage in the path to the MQSA's next reauthorization in 2007. But it is likely that the report's recommendations will carry weight with Congress.

Mikulski would like to see some of the proposals put into place sooner rather than later.

"I requested this study to gain expert recommendations on how we can strengthen this important program," she said. "I'm looking forward to working with the FDA on many of the IOM recommendations that can be implemented immediately."

Sure to promote controversy is the IOM's idea of expanding MQSA to other breast imaging modalities and renaming the legislation the Breast Imaging Quality Standards Act (BIQSA).

It's an idea whose time has come, according to Forman. Breast imaging has gone beyond the standard two-view mammogram that was common when the MQSA was introduced in the early 1990s. Stereotactic biopsy, breast MRI, and ultrasound are now routinely used, but the quality and accuracy of the techniques vary widely.

"We're not just talking about mammography anymore," he said. "We're talking about breast imaging, and that should be included in the overall regulation."

Others are less sure. Kolb would like to see the FDA obtain a better handle on mammography before it moves on to regulating other modalities. The problem, he said, is that neither ultrasound nor MRI is as easy to measure as mammography, in terms of technical quality.

"In ultrasound, baseline studies conducted prior to the ACRIN 6666 trial showed operator variability all over the map," he said. "We take a standard four views in mammography, which eliminates much of the variability from a technical standpoint. We know exactly what the view is to begin with. But there are no such criteria for ultrasound."

Pisano sees an advantage to the expansion to BIQSA. Such an umbrella rule would make it more difficult for breast surgeons and other nonradiologists to perform substandard breast imaging.

"Radiologists in general tend to perform imaging tests at the highest quality possible," she said. "Other specialists, in other disciplines, may be performing low-quality breast ultrasound and MRI. The positive side of agreeing to more regulation is that the best quality imaging is provided to patients."

Kopans admits to mixed feelings about the proposal. The existing rule has improved mammography, and other modalities might benefit as well. The key is making sure that regulations are intelligent and not too onerous, he said.

Mammography practice is not for the faint of heart. Logan-Young calls mammography the most difficult job in radiology, which is the reason why she chose it. The MQSA has not made mammography easier, but it has made it better. But there is plenty of room for improvement, according to Kopans.

"Aspects of MQSA need to be fixed, but we can build on it," he said. "It has to be done with radiologists as partners, not targets for corrective action. I'm impressed that radiologists are still hanging in there, reading mammograms. It's impossible not to miss some cancers every once in a while-no human being is perfect."

--

Ms. Dakins is feature editor of Diagnostic Imaging.

TABLE 1. INSTITUTE OF MEDICINE RECOMMENDATIONS FOR IMPROVING MAMMOGRAPHY QUALITY

Improve mammography interpretation

Revise and standardize the medical audit component of MQSA

Facilitate a voluntary advanced medical audit with feedback

Designate breast imaging centers of excellence and undertake demonstration projects and evaluations within them

Further study the effects of CME, reader volume, double reading, and CAD

Revise MQSA regulations, inspections, and enforcement

Modify regulations to clarify the intent and address current technology

Streamline inspections and strengthen enforcement

Ensure an adequate workforce for screening and diagnosis

Direct FDA to collect and analyze data on mammography workforce and services

Devise strategies to recruit and retain skilled breast imaging professionals

Make more effective use of breast imaging specialists

Improve breast imaging quality beyond mammography

Mandate accreditation for breast imaging methods beyond mammography, including ultrasound and MRI

Source: Improving Breast Imaging Quality Standards, Institute of Medicine, May 2005

TABLE 2. MQSA-CERTIFIED FACILITIES AND PROCEDURE VOLUME

Number of certified facilities : 8929

Number of accredited mammography devices : 13,640

Certified facilities with digital mammography devices : 607

Number of accredited digital mammography devices : 819

Total annual mammography procedures as of 4/05 : 13,486,163

Source: FDA, April 2005